Site Index | Site Tour

    or   Search Tips

National Center on Secondary Education and Transition: Creating opportunities for youth with disabilities to achieve successful futures.

State Contacts
Web Sites

Teleconference Transcript

Transcript of NCSET teleconference call held on February 17, 2005

The Disproportionate Representation of Culturally and Linguistically Diverse Students in Special Education


Elizabeth Kozleski, Ph.D., Co-Principal Investigator
National Center for Culturally Responsive Educational Systems (NCCRESt)

View the PowerPoint presentation referenced in this teleconference (in PDF format: 48 slides, 10.69 MB). (Also available as an accessible Word document.) Small numbers in the text below (i.e., 35) refer to the slide being discussed.

DR. SHARPE: Good afternoon and welcome to 1 "The Disproportionate Representation of Culturally and Linguistically Diverse Students in Special Education." I’m Mike Sharpe, director of the North Central Regional Resource Center (NCRRC), housed at the University of Minnesota. This teleconference is co-sponsored by the NCRRC in partnership with the National Center on Secondary Education and Transition and the OSEP Exiting TA Community of Practice.

The topic of disproportionality is of particular interest to many states, including states in the north central region. I’ve seen considerable evidence that states and districts are looking for effective means of addressing disproportionality in special education identification, and also the impact of disproportionality on transition-related outcomes, including graduation and dropout rates. Because of the interest of our members in disproportionality, the NCRRC, the Exiting TA Community of Practice, and the National Center on Secondary Education and Transition have developed a series of teleconferences on this important topic. Mark your calendars for the following dates:

  • March 31 at 1 p.m. Central Time, Anthony Sims of the Institute on Educational Leadership will present "Minority Disproportionality in Special Education and the Achievement Gap—Common Issues, Shared Solutions," and
  • April 14 at 2 p.m. Central Time, listeners will hear about the Wimberley Project, a project to reduce disproportionality in the Duval County, Florida public schools.

In addition, we are currently in the planning phase of presenting a state perspective on disproportionality as part of this teleconference series.

Dr. Chris Bremer of the National Center on Secondary Education and Transition will be today’s moderator.

DR. BREMER: Today we are pleased to have Dr. Elizabeth Kozleski, a professor and associate dean at the University of Colorado at Denver and Health Sciences Center. Her expertise is in the areas of systems change, inclusive education, and professional development in urban education. Her research interests include teaching, teacher learning and urban education, multicultural educational practices in the classroom, and the impact of professional development schools on student and teacher learning. She was a public school special education teacher for seven years before earning her doctorate from the University of Northern Colorado. Currently, she is co-principal investigator for the National Center for Culturally Responsive Educational Systems, or NCCRESt, and the National Institute for Urban School Improvement. Dr. Kozleski’s expertise in teacher education and urban education both support her work with the Council for Exceptional Children, Teacher Education Division; the American Association of Colleges for Teacher Education; Harvard’s Civil Rights Project; the Colorado Partnership for Educational Renewal; the National Center for Educational Outcomes; the American Institutes for Research, National Board for Professional Teaching Standards; and the Minnesota and Delaware Departments of Education, among others. Dr. Kozleski’s research and personnel preparation efforts have been funded by the U.S. Department of Education, Office of Special Education Programs; the National Education Association; and the Colorado Department of Education. She has presented her work at conferences in the U.S., Asia, and Europe, and has received awards for teaching, service, and research at her university.

DR. KOZLESKI: Good afternoon, everybody. I’m really glad to be here, and I am going to try as best as I can to convey what’s on the PowerPoint, if you don’t have an opportunity to pull it off the Web site. 2 I’m going to walk you through an agenda today that lays out a definition of disproportionality, the ways that we measure it, why we should pay attention to it, what we know about it, what should we do about it, what is NCCRESt doing, and how can we work together to improve outcomes for all students.

3 A 2002 report from the National Academy of Sciences on minority students in special education and gifted education defines disproportionality “from the enactment of the 1975 federal law requiring states to provide a free and appropriate education to all students with disabilities, children in some racial/ethnic groups have been identified for services in disproportionately large numbers.” The 2002 report was the second time that the National Academy of Sciences had taken a look at the issue of the over- and under-representation of children from culturally and linguistically diverse backgrounds in special education. Between the ’82 report and the 2002 report, there was very little change in the status of disproportionality. There were particular groups that were over-represented in 1982, and they continued to be over-represented in 2002.

4 Disproportionate representation is not a result of intrinsic or family-based deficits on the part of children. 5 What we think are contributors to disproportionality are policies at the local level, district level, and state level that encourage the identification and placement of children in special education without questioning the cultural profiles and cultural implications that children bring to the learning process. Another contributor may be the belief about what our teaching force brings to their classrooms in terms of preferences for the mainstream—the dominant culture perspectives of what constitutes achievement and success in the modern U.S. in terms of outcomes for adults and the practices that are in place in schools and classrooms that prefer certain kinds of behaviors and certain kinds of academic learning styles perhaps over others. It’s this combination that is particularly problematic in terms of policies, beliefs, and practices that together may create a context in which certain kids tend to get selected for special education services to a greater degree than other kinds of children.

Another issue around disproportionality is not only its presence in over- and under-identification in special education, but also a concomitant set of statistics in the general education environment that deal with things like, what’s the proportion of children from minority backgrounds who are taking SATs, who are in Advanced Placement classes, who are being given opportunities for service learning and job shadowing for career development and for entrance into high status professions in high school. It has both a general education and a special education component.

6 One of the things that we don’t know a great deal about from a research perspective are the intersections of the learning research, the research on disability, and the research on the sociocultural nature of inclusion in community groups. We need to answer what all three of these things together do and create for individual kids in classrooms when we try to begin to look at what kind of classroom curriculum or classroom instruction may be preferential. We also need to begin to look at research that has sizeable groups of children who come from non-majority cultures in the groups who are receiving treatment and non-treatment, so we can begin to understand which of the kinds of curriculum, instruction, and school supports are necessary in order to support kids from culturally and linguistically diverse backgrounds.

7 The background assumptions that we make about learning and development are that there are individual contributing factors to how a child might be successful in school. 8 There are also contextual factors that are created by the institutions themselves, and those interact in many different ways with the family status of a particular child, the community in which a child finds themselves, the school, and then the social pressures that are placed on schools in terms of being successful.

Disproportionality is a very complex issue nested inside multiple dimensions of the problem. If we deal with disproportionality at only one level of the complexity, we will solve problems locally but we won’t be able to solve problems at the system level.

The National Center for Culturally Responsive Educational Systems is interested in creating systems responses to the issue of disproportionality. 9 In order to do that, we have to talk about how we measure disproportionality. 10 The current best practice around measuring disproportionality is using something called the Relative Risk Ratio, where we compare the risk of identification for one group of students with a particular ethnic background. For instance, students who are identified as African American, or black, compared to the risk for white students or for all other students. The way we do that is to use two risk indexes, the relationship between all of the black students in a school population and the subset of students who are black and labeled for potentially mental retardation or learning disabilities or emotional disturbance, and we get a risk index for that particular group. Then we compare that risk index to the risk index for white students by taking the population of all the white students and leveraging it against the population of white students who have been labeled for mental retardation, and we compare the two risk indexes. We then get a Relative Risk Index. When that risk index rises above 1, we’re beginning to look at an issue around over-representation. When the risk index gets to 1.2 or above, that’s a place where a school or a district or a state might want to begin to look at that particular issue.

11 The 2002 National Academy of Sciences report asks, “If IDEA provides extra resources and the right to a more individualized educational program, why would one consider the disproportionate representation of minority children a problem?” 12 They answered, “in order to be eligible for the additional resources a child must be labeled as having a disability, a label that signals substandard performance. And while that label is intended to bring additional supports, it may also bring lowered expectations on the part of teachers, other children, and the identified student. When a child cannot learn without the additional supports, and when the supports improve outcomes for the child, that trade-off may well be worth making. But because there is a trade-off, both the need and the benefit should be established before the label and the cost are imposed.” 13 When we consider the issue of disproportionality, we need to consider:

  • Special education may not provide the supports that a student needs
  • The disability label may stigmatize a student as inferior
  • Results in lowered expectations
  • Potentially separates a student from peers
  • May lead to poor educational and life outcomes
  • Students may be denied access to the general education curriculum
  • May result in dropout
  • Students may be misunderstood or underserved in general education

14, 15 NCCRESt offers a lot of information about the last four years of data that states have provided to the federal government around disproportionality. 16 The map of the U.S. on the NCCRESt Web site shows all of the disability categories for African American students, the lowest risk ratio to the highest risk ratio, how states compare to one another in terms of the risk for African American students to be labeled in any disability category, and what that risk might be. The western states seem to be much hotter spots for all disabilities than the eastern side which are predominantly blue states, with a few exceptions in North and South Carolina.

17 If you go to the next map, the color suddenly reverses. Now the western half of the U.S. is mostly blue, and the eastern half of the U.S., particularly in the southeastern corridor, is predominantly red. That’s because we asked a different question in this map. In this map, what we’re looking at is the risk ratio in 2002-03 for African American students to be labeled for mental retardation. What’s happening is that maps are recalculating the risk ratio every time we ask a different question. We can encourage states to actually look at this data in terms of the risk ratio for different populations of kids in their states. The second map asked the risk ratio versus other races for students who were African American who were labeled for mental retardation. 18 The third map looks at the risk ratio for students who are labeled for mental retardation who were African American, but the risk ratio is comparing African Americans to whites in this case.

As the denominator in our risk ratio changes, the maps change color; as the disability category changes, the maps change color; as the ethnicity of the kids changes, the maps change color. You begin to see patterns and shifts in what’s going on in different states. This allows people who are working at the policy level to ask questions about the policies in those states that are creating particular patterns, and we can begin to look at the kinds of policies that can encourage the most equitable participation in special education.

19 I’m now moving onto a slide that shows data series and data groups that you can actually ask for. We’ve grouped all of the states according to their ethnic composition. Iowa, Maine, New Hampshire, Vermont, and West Virginia are grouped as homogeneous whites because the preponderance of their population has identified themselves as white in that state. To predominantly multi-racial states, like California, Florida, Illinois, Nevada, New Jersey, New Mexico, New York, and Texas there is no one race in that state that is a majority.

20 By organizing states into predominantly bi-racial states and moderately multi-racial states we can ask questions about the risk ratio change when we look at states who fall into these different categories, and then we can move on and begin to look at a set of trend graphs that show what happens when we pick particular groups or clusters of states. I’m now looking at the trend for the identification of African American students in high incidence disabilities—emotional disturbance, learning disability, and mental retardation—for states that are predominantly bi-racial, including Delaware, Alabama, and Maryland over the past four years. In Delaware the risk ratio for African Americans to be identified in high-incidence disabilities went from 2.59 to 2.53. They have a relatively stable trend line showing that students who are African American are more than twice as likely than students of other races to be identified for high-incidence disabilities. In Alabama the trend line increases slightly from 1.6 in 1999 to 1.73 in 2003, and Maryland’s trend line increases moderately from 1.57 to 1.60.

21 The next trend map shows states that are predominantly multi-racial including New Jersey, whose trend line moved from 1.41 to 1.49 over a 4-year period; Florida, whose trend line decreased from 1.61 to 1.57; and California, which started at 2.07, went down to 2.01, and then went back up to 2.14 in 2002. Multi-racial states seem to be having difficulty working in high-incidence disabilities in terms of their identification of students who are African American.

22 The third trend graph looks at predominantly homogeneous white states including Maine, West Virginia, and Iowa. Iowa is hovering at about the 2.4 level, 2.4 times as likely to identify students who are African American in high-incidence disabilities. Maine is at 1.4 and West Virginia at 1.23 in 2003. Maine’s number went down significantly one year, and then started to creep back up.

23 This interactive Web site shows states in relationship to each other across many different kinds of disability categories across different ethnic groups. It also shows trend data and data points by state, and the data series in terms of risk ratio, and—24 look at bar chart—disproportionality by race and disability. The bar chart compares California, Connecticut, Delaware, and Iowa for students who are African American and high-incidence disabilities for the 2002/03 school year. Three of those states are more than twice as likely to identify students who are African American as having high-incidence disabilities.

We all know that states are an amalgam of what local education agencies actually do with their students. One of the things that we’ve also done at NCCRESt is to try and ask the question, what happens when we actually look at the same kinds of patterns at a local education agency level. 25 Another graph is based on the snapshots of data from Clark County in Nevada, which is the Las Vegas school system, the fastest growing school system in the country; from Denver, Colorado; from Miami, Florida; from Washington, DC; and from Chicago, Illinois. We analyzed the data across the five school systems, and we noticed that in any given school, where there is a population that is a minority, whether the population is white or Hispanic or African American or Asian, and there is a dominant majority population in that building, there is a high risk for the minority population in that school to be over-identified in special education. The table shows us that where population distributions are heavily weighted in terms of one large majority and a very small minority, you get a lot more disproportionality or disproportionate identification of students for special education.

26 The next map shows Chicago, and it’s organized in a way that allows us to look at schools that have census poverty levels that are greater than 20% and schools with poverty levels that are less than 21%. It maps the hotspots of Chicago where there seem to be sets of schools that are over-identifying their students who are African American in particular clusters. Those hotspots where the disproportionate risk ratio seems to exist for African Americans, we look at that in relationship to whether or not the neighborhoods and the schools that are implicated are high poverty or not, there seems to be very little relationship between poverty and the disproportionate identification of African American students. This kind of a map also appears for the other cities that we’ve been mapping. The idea that it’s poverty as well as race that seems to be playing out in terms of which children are getting identified, at least in the cities that we’re taking a look at, seems to not be as prevalent as the data expected that it might.

27 To view your own situation, at least at the state level, go to the NCCRESt data manager site at, click on the login button, click ‘enter the campus now,’ and then type in the username ‘nccrest learner’ and the password ‘nccrest,’ and then click ‘data maps’ and you will be in our data map manager, and you can play in there and take a look at what’s going on for your particular state.

28 What should we do about disproportionality? 29 One of the things that seems to be important to states is “Eliminating disproportionality is an adult issue.” That’s a quote from the Superintendent of the Seattle schools. “We must change the way we think about ability, competence, and success and encourage schools to redefine support so that the need to sort children is reduced.” That’s testimony before the President’s Commission on Special Education in 2002.

30 To provide technical assistance and dissemination to states in the search for solutions to their widespread underachievement and disproportionate placements in special education experienced by culturally and linguistically diverse students, NCCRESt has organized work into four key areas. One of them, continuous improvement, is only possible if we provide people with the data they need to help them understand the nature of the problem in their particular situation—the role of the data manager maps.

We also believe that we need to synthesize the literature, from intervention literature to socio-cultural literature to literature that helps us understand what culturally responsive pedagogy might be, so that we can put tools in the hands of practitioners so that they understand things that they might think about, practices that they might institute in their classes, and assessment systems that they might integrate into thinking about the nature of student needs and what that means for instruction.

Our Web site now has eight practitioner briefs to which we are continuing to expand by adding a series of research-based articles in a variety of general education and special education journals. People need technical assistance and professional development to make the shift in their practice, so we have a whole strand at NCCRESt that deals with the development of leadership academies around culturally responsive educational practices.

Finally, we need to get the word out as much as we possibly can, to network with other people around the country, like Anthony Sims, who is going to be on the next teleconference, who are doing work around disproportionality so that we can become more powerful together in addressing this issue. We think it’s a systemic issue, and we need to think systemically about solutions. 31 In any educational system we have people, policies, and practices that need to be accounted for and we cannot make progress unless we operate in multiple arenas simultaneously.

32 The culture and language and heritage of all students and families are valued, respected, and used to facilitate learning and development. That is an essential feature of any culturally responsive system. 33 Features of culturally responsive educational systems include practitioners and administrators who assume responsibility for the learning of all students from all cultural and linguistic backgrounds; systems where every student benefits academically, socioculturally, and linguistically; and systems where all students have access to high quality teachers, programs, curricula, and resources. 34 If we think about this as a systemic change initiative, and we think about this as simultaneous renewal at multiple levels of the system, 35 we’ve got to think about whether or not, at the local school level and local classrooms, every child has access to the curriculum, whether or not every child is able to participate in the curriculum, and whether or not every child has equity in terms of the instruction and the feedback and the materials that are made available for that student to learn.

36, 37 Building culturally responsive systems requires engaging people, and one of the things that we’ve been doing a lot of work around is focusing on making sure that the voices of all of the students in a building and their families are present, are encouraged to participate and to offer thoughts that may be unique or not expected, and let them build a community of practice in their own buildings that supports the kinds of traditions and heritages and assets that those families bring into the school. 38 We want to encourage people at the school level, the district level, the state level, and the federal level to examine policies, to ask whether or not the policies educate and move the agenda forward, whether they help to inform the practice, whether they’re equitable in terms of their distribution of resources and opportunities, whether they emancipate, and whether or not they create access for all students and families. 39 To do that it requires conversations like the one that we’re having today and a set of tools.

40 What should be the focus of change? It needs to be at the classroom, school, district, state, and federal levels. 41 Connecting the dots, the data, and change requires robust and sustainable change, requiring masterful use of evidence. Effective school improvement can only exist within a context of clear information; clear, specific goals and outcome measures; and information systems that provide just-in-time information.

42 Our technical assistance strategy is to help states build capacity around their professional development and technical assistance, 43 to build skill sets at the state education agency level, and to work in teams that cut across the state education agency, advocacy, and district personnel so that we make sure that within any given state the specific issues that are leading to disproportionality in LEAs are the focal point for the kinds of activities—technical assistance and professional development—that are going on to make a change.

We identify districts that are improving, that can serve as beacons for other districts, and build strategies to support their work, and then we identify districts that need to learn, because they’re in the development phase, so that we can build their capacity.

There are many assumptions in the work we’re doing—that this is not a special education problem, but one that deals with the disproportionality issue requiring a focus on education as it’s developed in classrooms and schools across the population of kids, as opposed to working on it from within the special education milieu.

44 We do a lot of different kinds of things at NCCRESt to facilitate this and we encourage you to participate in those kinds of activities, to visit our Web site, to get on our eNews subscription list, to continue to participate in communities of practice like this in partnerships with technical assistance and RRC centers. In September 2005, we are in the process of developing a national forum on disproportionality that will be an open invitation to people around the country to come. In addition, we are working very intensively with nine states to develop models at the state level for solving the issues of disproportionality.

45 Working together to make a difference includes understanding the assumption that student characteristics are both psychological and sociocultural, and that they bring both of those traditions, those avenues to their learning that curriculum must address. Teachers need to understand how to work in culturally responsive ways, and schools need to attend and respond to patterns of performance, selection, and inclusion in their buildings. Inclusiveness in that we have not only students with disabilities but also students who come from many different cultural and ethnic backgrounds who we ensure are included in all of our programs. There’s also an acknowledgement that this is difficult work and technical assistance needs to be geared to help schools do this—to understand how to ask the questions, how to look at their data, and how to understand the kinds of skills they need to develop in their teachers to make this work successful.

46 Keep in mind that there are factors that influence the state and district capacity to do this work: the policy environment; the allocation of resources; the availability of state personnel; the integration of information systems; the linking among organizations, programs, and projects. 47 If we’re going to change the way that things are, we need to understand the data, we need to focus on the classroom, we need to use whole-school improvement models, we need to build networks of schools, and we need to do policy review, training, and reform in order to change the current context.

48 This presentation looked at what disproportionality is, how we measure it, how we should pay attention to it, what we know about it, what we should do about it, what NCCRESt is doing about it, and potentially some activities that we could work on together to improve outcomes for all students.

PARTICIPANT: I’d like to ask who the nine states are that you are working with?

DR. KOZLESKI: Connecticut, New Jersey, North Carolina, Tennessee, Louisiana, Ohio, Wisconsin, Iowa, and one more. Why don’t we talk about something else while I remember them?

PARTICIPANT: You mentioned the importance of knowing how to read the data. Are there particular dangers that the data might, or typically in this field, lead people astray?

DR. KOZLESKI: I think the biggest point to make about looking at the data around disproportionality is to understand that if we work just to change the data, we might work on the wrong thing. A really quick fix to solving the problem is just simply to stop identifying children of color. That may or may not be the right thing to do. That thinking about what the data means, and then the response to the data needs to be a much deeper conversation than just making the data the way that they should be, or the assumption that they should be, which is kids of all different races get identified at exactly the same levels. The data may make us choose fixes that may not really be deep solutions to the issue.

PARTICIPANT: What is the definition of disproportionality and do we automatically assume from the presentation that when you have a higher proportion of a particular race or gender or ethnic group, that is automatically a problem? Do you not start with the assumption that it’s somewhat problematic?

DR. KOZLESKI: To answer the first question, I gave you the definition from the National Academy of Sciences. I did not give you numerical or technical definitions of disproportionality because then we are dealing with numbers as opposed to the root cause.

To answer the second question, you don’t know until you look deeper into the situation what you’re looking at, what is going on, what are the features of the particular context that you’re looking at, how is it that certain sets of kids are coming into special education, what is the referral process, what are the teaching technologies that are being used in a general education classroom when the kids are being identified for special education, and what are the preventative measures that are being used? There is a whole set of questions that need to be asked of the data. Simply create the opportunity to ask those questions.

PARTICIPANT: Once disproportionality is identified, as you said, it may or may not be a problem in and of itself—it’s indicative of, perhaps, other problems. Have you found in the states and districts that you’re working with that there seem to be other measurable factors closely associated with it, like poor performance on district or statewide tests or greater incidence of dropping out or fewer graduates? I think you mentioned participation in SAT tests, those sorts of things?

DR. KOZLESKI: We are looking at a district right now in California where we are looking at all of those measures together and looking at what the relationship is between not only the negative factors—suspensions, expulsions, dropout, lower achievement rates—but also the other side of the equation, which is access to Advanced Placement classes, and looking at the relationship and the predictive value for disproportionate representation.

PARTICIPANT: I’m from Lawrence, Kansas, at University of Kansas. What is the prevalence of minority children with Asperger’s/Autism, where can I get that information, and can you use data maps to get the answers?

DR. KOZLESKI: Our data maps only map the disability categories that the feds use, because our data comes directly from OSEP, from Westat. We don’t track children who are labeled as autistic or Asperger’s Syndrome by themselves.

MR. McCAIM: Could the speaker talk a little bit about under-representation of ELLs in special education? In Oregon we have a small representation of it, lower than national norms for English Language Learners being referred to special education. Dr. Baca was here recently and this was a new phenomenon to him as well. I want to know if the speaker could comment on that?

DR. KOZLESKI: If you have a chance to go visit our data maps on the NCCRESt Web site, you can look at under-representation issues, particularly for students who come from Hispanic categories. We know that not all children who are identified as Hispanic speak either Spanish or English as their first language, so it’s not a perfect match. That is an issue that is appearing in more than Oregon as a phenomenon to worry about.

MS. AUTIN: This is Diana Autin from the Region 1 Parent Technical Assistance Center and the Statewide Parent Advocacy Network of New Jersey. Parents play a critical role in helping avoid the over-classification of their children, if their children are classified, and the over-segregation of their children. I’m wondering if the work in the states that you’re working with involves working with African American parents in terms of building their capacity to make good decisions about whether or not their children should be classified and if so, ensuring more inclusive placements for their children?

DR. KOZLESKI: Well, we agree with you that families are a real core piece of this. We have a project that is going on here in Denver with the Latino Statewide Family Organization that is actually working with families through a grassroots organization to teach them about the issues of disproportionality and to teach them about ways to think about the conversations that they might have with teachers and child study teams in their buildings around that very issue. We’re hoping to develop a process with them that we can build with other states dealing with other ethnic groups around that particular issue.

DR. BREMER: I’d like to thank Elizabeth for sharing her time and expertise with us. If you’re interested in learning more about transition issues and helping youth graduate and achieve successful post school outcomes, we invite you to join the OSEP Exiting TA Community of Practice at I know Elizabeth will invite you to continue to visit the NCCRESt Web site for all the wonderful resources there. You might also check out the NCSET Web site for recent postings. The next NCSET Exiting TA Community teleconference is scheduled for Tuesday, February 22. The presenter will be Dr. Martha Thurlow and the topic is the implications of standards, assessments, and accountability on graduation requirements and diploma options. We hope to see you all then. Thank you for joining us and have a great day.



E-mail this page
Download PDF

^ Top of Page ^

Publications  |  Topics  |  E-News  |  Events  | State Contacts

Web Sites  |  About NCSET  |  Home  |  Search

National Center on Secondary Education and Transition
Institute on Community Integration
University of Minnesota
6 Pattee Hall
150 Pillsbury Drive SE
Minneapolis MN 55455

© 2001-2017 Regents of the University of Minnesota
The University of Minnesota is an equal opportunity educator and employer.
Online Privacy Policy

This page was last updated on November 29, 2017.